acetate or 1-hexene may also be used for the synthesis of aro-
matic carbonyl compounds when lower reaction rates are
acceptable. For example, using these two oleÐns for the syn-
thesis of acetophenone, 97 and 92% yields respectively, were
achieved in 5 h, and the only byproduct was ethylbenzene (for
conditions see Table 3).
The high reactivity of cyclohexene as a hydrogen acceptor
may be astonishing at Ðrst sight, as cyclohexene is widely used
as an H-donor in transfer hydrogenation reactions over sup-
ported Pd catalysts.22h25 In fact, when reÑuxing an
acetophenone : 1-phenylethanol \ 94 : 6 mixture in cyclo-
hexane solvent over Pd/Al O , the ketone : alcohol ratio
Scheme 3 Byproduct formation via isomerization over Pd during
dehydrogenation of allylic alcohols.
2
3
increased or decreased depending on the cyclohexene concen-
tration. When the alcohol : cyclohexene ratio was very low,
the side reaction depicted in Scheme 1. Another, dominant
route may be isomerization via the half-dehydrogenated state
as shown in Scheme 3.20 For example, dehydrogenation of
cinnamyl alcohol (12) a†orded a mixture of 3-phenyl-1-propa-
nol, 3-phenyl-1-propanal and cinnamaldehyde as major pro-
ducts.
The low reactivity of aliphatic (14) and cycloaliphatic (15)
alcohols can be exploited for the selective synthesis of aro-
matic ketones. In the example shown in Scheme 4 a reactant
possessing both aromatic and aliphatic OH groups is mim-
icked by an equimolar mixture of 1 and 14. As expected on
the basis of Table 3, competitive dehydrogenation of 1 and 14
a†orded in 3 h higher than 90% yield to acetophenone while
1
: 40, acetophenone was slowly hydrogenated to 1-
phenylethanol (0.5È1% conversion per hour). The double role
of cyclohexene can explain our observation that increasing the
excess of cyclohexene in transfer dehydrogenation reactions
enhanced the initial rate of alcohol conversion but the Ðnal
yield could not be improved.
We can conclude that transfer dehydrogenation of alcohols
with the Pd/Al O Ècyclohexene system o†ers a facile and con-
2
3
venient synthesis of aromatic ketones. The amount of hydro-
carbon formed by hydrogenolysis of the CHÈOH function can
be minimized by selective poisoning with trace amounts of a
tertiary amine, such as Et N.
3
1
4 remained predominantly unattacked. For comparison, the
Pt metal catalyzed aerobic oxidation of aliphatic alcohols is
not slower than that of aromatic alcohols. As an example, oxi-
Experimental
dation of 1-heptanol and 1-dodecanol on PtO in heptane at
2
6
0 ¡C a†orded 26 and 77% aldehyde yields in 1 and 0.25 h,
Materials
respectively.21 Under the same conditions, oxidation of 10 to
benzaldehyde a†orded 78% yield in 1 h.
Five wt% Pd/C (Fluka, product no. 75992), 5 wt% Pd/Al O
2 3
(
(
Johnson Matthey, product no. 56482), 5 wt% Pd/CaCO
3
5 wt% Pd/TiO (Johnson
Fluka, product no. 76032),
wt% Pd/organopolysiloxane (Degussa,
2
Matthey) and
5
Discussion
product no. 1994-8874) catalysts were used for transfer dehy-
drogenation without any treatment. 1-Phenylethanol (Aldrich,
The Pt metal-catalyzed aqueous phase oxidative dehydroge-
nation of alcohols with molecular oxygen a†ords high yields
to ketones and carboxylic acids under mild conditions. A
drawback of the method is the high activity of Pt and Pd-
based catalysts in the oxidation of almost all types of primary
and secondary alcohols. In contrast, transfer dehydrogenation
over supported Pd catalysts using cyclohexene as the hydro-
gen acceptor allows fast and selective synthesis of aromatic
ketones only; aliphatic and cycloaliphatic primary and sec-
ondary alcohols are barely reactive under the conditions
applied. Application of cyclohexene in reÑuxing cyclohexane is
convenient as the coproduct formed from the hydrogen accep-
tor is identical to the solvent.
[
98%),
DL-6-methoxy-a-methyl-2-naphthalenemethanol
(
(
Acros, 98%), benzyl alcohol (Fluka, 98%), cinnamyl alcohol
Fluka, 97%), 1-dimethylamino-2-propanol (Fluka, 98%), 1-
octanol (Fluka, 99.5%), benzhydrol (Aldrich, 99%), benzoin
Aldrich, 98%), 3,5,5-trimethyl-2-cyclohexene-1-ol (Fluka,
5%), 1-(4-chlorophenyl)ethanol (Aldrich, 98%), 1-(1-naph-
(
9
thyl)ethanol (Fluka, 99%), a-pyridoin (Aldrich, 99%), 1-(4-
methoxyphenyl)ethanol (Aldrich, 99%), ethylmandelate
(Merck, 98%), 1-(4-methylphenyl)ethanol (Lancaster, 97%),
cyclohexanol (Merck, [99%) were used as received. 1-
Indanol (Merck, [98%) was puriÐed by sublimation. Analyti-
cal grade solvents were used as received. Cyclohexene (Fluka,
9
9.5% or Merck, [99%), 1-hexene (Fluka, 98%), methyl acry-
Several liquid oleÐns have been tested together with
Pd/Al O and the conditions have been optimized for cyclo-
late (Merck, [99%), cyclopentene (Aldrich, 99%), styrene
2
3
(
Aldrich, 99%), methyl vinyl ketone (Merck, [95%) and vinyl
hexene, which was the most reactive hydrogen acceptor (Table
). We have to emphasize that other oleÐns such as vinyl
acetate (Aldrich, 99%) were used as H-acceptors. Tri-
ethylamine was a Siegfried Synopharm product (99%).
1
Methods
Catalyst (0.10 g), alcohol (1.0 g), hydrogen acceptor, 30 ml
solvent and Et N (0.01È0.02 g), when used as a catalyst
3
poison, were put in a 100 ml glass reactor. Air was replaced by
Ar and the reactor was put into a preheated oil bath. The
reactions were carried out at reÑux temperature with intensive
magnetic stirring. Alcohol and cyclohexene conversion and
the product distribution were determined by GC analysis
(Thermo Quest Trace 2000, equipped with an HP-FFAP cap-
illary column and FID detector). Yields in the dehydroge-
nation of 8 and 9 were determined by 1H NMR (Bruker
Scheme 4 Competitive transfer dehydrogenation of 1 and 14. Condi-
tions: 0.10 wt% Pd/Al O , 0.5 of each alcohol,
alcohol : cyclohexene mol ratio \ 1 : 2, 30 mol cyclohexane, reÑux
80 ¡C).
Advance DPX300, 300 MHz, in CDCl ). Products were iden-
g
5
g
3
2
3
tiÐed by GC-MS, NMR, and GC analysis of authentic
(
samples.
1166
New J. Chem., 2001, 25, 1163È1167